Wikipedia talk:Manual of Style/Dates and numbers

AD Vs CE Discussion

Discussion moved from Wikipedia talk:WikiProject Military history/Coordinators/Strategy think tank#AD Vs CE by Macarenses. Note that there were a few comments there that were not moved with the rest of the discussion. Dana boomer (talk) 15:32, 21 December 2010 (UTC)

I believe that the date style in the article within the scope of the project should be changed from the Christian centered AD-BC system to the BCE ireligious system. Since both systems denote the same years it is just a matter of changing the acronyms in every article. A small change in practice but a very important one in that this way the article doesn't take a religious stance. I mean can you imagine if instead of the seemingly harmless "AD" we would actually write the true meaning of the acronym- "the battle was fought in the year of OUR LORD 1627"- I think not! and yet we seem satisfied to keep this religious note in copious amounts, littering even articles regarding wars fought hundreds of years before "the anointed one".Thx for hearing me out--Macarenses (talk) 14:27, 21 December 2010 (UTC)

WP:ERA states:

• AD and BC are the traditional ways of referring to these eras. CE and BCE are becoming more common in academic and some religious writing. No preference is given to either style.

• Do not use CE or AD unless the date would be ambiguous without it (e.g. "The Norman Conquest took place in 1066" not 1066 CE or AD 1066).
• BCE and CE or BC and AD are written in upper case, unspaced, without periods (full stops), and separated from the year number by a space or non-breaking space (5 BC, not 5BC).
• Use either the BC-AD or the BCE-CE notation, but not both in the same article. AD may appear before or after a year (AD 106, 106 AD); the other abbreviations appear after (106 CE, 3700 BCE, 3700 BC).
• Do not change from one style to another unless there is substantial reason for the change, and consensus for the change with other editors.

The reason it opts for whatever the status quo is for that article as opposed to choosing one system for the whole of Wikipedia is that there's no consensus on which to use. I don't think WP:MILHIST should be used to usurp that. Years ago there was an enormous debate on the issue which didn't lead really anywhere. As this project is one of Wikipedia's largest, I think we have a responsibility to tread carefully. If this project adopts one over the other it will be pointed to by whichever side is chosen as giving them legitimacy. Both AD/BC and CE/BCE are used in academic literature; usage should be decided on a case-by-case basis and I feel that an attempt to enforce one style over another universally would be a waste of time at best. Nev1 (talk) 14:38, 21 December 2010 (UTC)

I get your point but the thing is that the current situation isn't neutral, not in the least. The fact that AD is used in academic literature doesn't seem to me to add that much weight to the argument as I've read academic publishings of Israeli scholars who write (translation) "...in the 1943 year of the Christian count.." and "...in the year 1245 according to their count.."- "their" being obviously Christians. Should we then allow for that format as well?. I believe a single non-religious format should be used. And if not then why shouldn't instead of "AD" to use the complete "Anno Domini" from time to time, it is what is means. "AD" in my mind cannot be viewed as anything but a Christian way and wikipedia should not be conscribed to the religious notion that every passing year is the year of jesus.--Macarenses (talk) 15:04, 21 December 2010 (UTC)

(via Wikipedia_talk:WikiProject_Military_history/Coordinators/Strategy_think_tank#AD_Vs_CE). I'm very suspicious of changes that seem to be motivated by a belief that it's not enough that we try to be a high-quality source of verifiable information but that we should also present that information in a way that conforms to some group or other's concept of "correctness"—whether that's because a common and widely-understood dating convention happens to have an archaic religious origin or because an article about the male reproductive organ has a picture of said organ. Realistically, even if we tried we'd be setting ourselves up to fail. While some might find BC/AD problematic, I'd be willing to bet an equal number would say the same of BCE/CE.

I also think perhaps you misunderstand what "neutrality" means in Wikipedia terms. In essence it means Wikipedia doesn't have an opinion... that is, we accurately reflect the sources used without imposing our own views on them. Your suggestion would seem to me to be violating WP:NPOV by changing dates that use BC/AD in the sources to CE/BCE precisely because of your personal interpretation. EyeSerene^talk 15:50, 21 December 2010 (UTC)

I have no personal preference in this regard, but I think that maintaining the system whereby articles choose a style on a case-by-case basis (using the current MOS guidance) is probably the best solution. I don't think a project-wide consensus is possible in regards to this issue, however, as per how we deal with the English variation issue, if editors use common sense and work together the goal of consistency within individual articles should be able to be achieved. AustralianRupert (talk) 21:29, 21 December 2010 (UTC)

Recent previous discussion Art LaPella (talk) 23:11, 21 December 2010 (UTC)

• While we're at it, I'm extremely offended by the use of Wednesday on Wikipedia, so I recommend we rename that to Third day (not "Fourth day", of course, because that would be basing things on the Judeo-Christian week sequence). But seriously, this has been discussed ad infinitum, and I myself brought it up less than 2 months ago, and the discussion went nowhere. The current compromise is the best we're going to get. The opinion that "CE/BCE" are neutral is laughable. The era is still based on a 6th-century monk's estimate of Jesus' birth year. Until we make a new calendar with a more secular epoch, this one will have a Christian bias. And what's wrong with that? As I've pointed out, our days of the week and months are named after Roman deities, so what's the blasted difference? It baffles me how some atheists (I myself am an atheist, but I see you're one as well) can be offended by such harmless mythology, and seemingly only when it's Christian mythology. Do you have a problem with Christianity specifically? If so, that bias is not Wikipedia's concern. As has been noted, AD/BC are just as much if not more notable than BCE/CE, so we are absolutely justified in using both systems, or even just AD/BC. — CIS ^(talk | stalk) 23:17, 21 December 2010 (UTC)

I propose that Wikipedia adopts the French Republican Calendar as a secular alternative. Nev1 (talk) 00:31, 22 December 2010 (UTC)

Tempting though it might be to prescribe consistency across Wikipedia about BC-AD/BCE-CE, there just isn't the consensus for it. As consensus is impossible, compromise is essential. I think the present guidelines are sensible. Michael Glass (talk) 00:11, 22 December 2010 (UTC)

• I’m not advocating a change to MOSNUM in this regard since editors will dig in their feet if anyone but God herself tries to tell them what to do in this regard. Since there is no central, all-powerful overseer of our style guides, Wikipedia is pretty much a state of regular chaos where one article uses *this* practice and another uses *that*.

  But, personally, I encourage editors to not be tempted to use Wikipedia as a platform in the vain hope of promoting the adoption of new-age ways of doing things; not until the practices become most common in the real world. Some wide-eyed wikipedians specializing in computer-releated articles tried in vain for three long years to promote the adoption of the IEC prefixes (mebibytes instead of megabytes) by flitting about and changing hundreds of articles nearly overnight. The result was some articles that spoke of “256 MiB of RAM” and still others that spoke of “256 MB of RAM”. It was pretty much an advertisement for how Wikipedia can do dumb things at times. Unfortunately, the naive effort didn’t help the world adopt the IEC prefixes one iota. All the IEC proponents accomplished was the inconsistent use of a writing style that drew untoward attention to itself and caused unnecessary confusion for our readership (both being highly verboten in all good technical writing).

  With very rare exception, you will only see “BCE” in writing; it is very seldom used in narrated form in TV, radio, and movies. Next time you see a TV documentary on Egyptian pyramids, note whether you hear the narrator speak of “Bee See” or “Bee See Eee”. I personally think is superior to use a writing style that reflects what people are accustomed to hearing. For those here who think that several dozen wikipedians can Make A Difference®™© in whether a new writing style catches on and attains widespread acceptance, think again; the IEC prefixes showed that notion to be fallacious. Wikipedia does best when it simply goes with the flow. Greg L (talk) 01:54, 27 December 2010 (UTC)

  • Chicago 9.35 uses the rare words (for Chicago) "personal preference"; AP Stylebook says to use AD/BC. (Chicago tries to keep academics happy when possible; AP doesn't.) I agree with pretty much all the replies, and well put. I just want to add that the trend is not to say "AD" at all, and the religious connotation probably has something to do with that. Some academics write BC but sprinkle their lectures with "before the common era" to make it clear that that's how they interpret the letters. - Dank (push to talk) 14:30, 29 December 2010 (UTC)
    • Absolutely right Sir! The whole thing is an amusing insight into political correctness, but it is Wikipedia's duty to follow rather than lead in these matters, and while a consistent and clear choice would be preferable, since we can't achieve it to date, the "leave well alone" principle applies almost as well. Rich Farmbrough, 15:03, 29 December 2010 (UTC).
      

• The obvious answer is to use stardate unless that upsets the Borg Jim Sweeney (talk) 09:12, 1 February 2011 (UTC)

Symbol for revolutions per minute

I've seen two symbols on Wikipedia for revolutions per minute:

• rpm
• RPM

The MOSNUM symbol for pounds per square inch is lower case 'psi'. Although MOSNUM doesn't yet state the symbol for miles per gallon, the convert template does show it in lower case. I propose that the table in MOSNUM is updated to show lower case 'rpm'. Any comments? Lightmouse (talk) 10:29, 14 January 2011 (UTC)

Agree Roger (talk) 12:03, 14 January 2011 (UTC)

Agree (unless ISO 80000 dictates otherwise). This would be in keeping with the SI convention of using upper case letters where people's names are concerned and lower case elsewhere except where there could be confusion (eg "T" for "teslas", but "t" for "tonnes", but litres may be "L" or "l" as "l" and "1" are easily confused). Martinvl (talk) 12:54, 14 January 2011 (UTC)

Agree, it's better if things are consistent. psi & mpg, thus rpm. Headbomb {talk / contribs / physics / books} 01:57, 15 January 2011 (UTC)

My caveat (which doesn't apply to three consecutive consonants like mpg or rpm) is that if you take away both full-stops/periods and capitalization or italicization, the abbreviation can sometimes look too much like a common English word, e.g. the Greek letter psi, the musical note mi, the exclamation ha [hectare is never so abbreviated in non-specialist US prose, which means American readers have absolutely no idea what ha might be] or the informal pa (one's father) for per annum. [Some commoner examples, which I can't recall off-hand, do throw me momentarily in reading British periodicals.] This is one of those many areas where a convention that is universal, familiar and perfectly-sensible in a scientific, engineering or technical article can confuse the reader of a non-technical page.—— Shakescene (talk) 18:57, 15 January 2011 (UTC)

Calories

A calorie (symbol cal) refers to a gram calorie while the kilocalorie (symbol kcal) refers to the kilogram calorie (also known as small calorie and large calorie respectively). When used in a nutrition related article, use the kilocalorie as the primary unit. In US-related articles, use the synonym dietary calorie with a one-time link to kilogram calorie.

• No, "kilocalorie" refers to 1,000 calories. Of course, 1,000 gram calories is equivalent to one kilogram calorie but semantically this is different statement.
• Watch the use of articles (a, an & the). We're talking about words here not units.
• Taken literally MOSNUM says we should prefer the kilocalorie over the kilojoule in nutrition-related articles (note the hyphen). Surely we should prefer the SI unit where practical. I think the intention was to prefer the (gram) kilocalorie over the (gram) calorie or the (kilogram) calorie (i.e. omitting the "gram"/"kilogram").

JIMp _talk·cont 21:27, 18 January 2011 (UTC)

That's because we should indeed prefer the kilocalorie - should we consider spelling it Calorie? - we should follow the usage of reliable sources, because as here, calories are what the reader is likely to understand and have seen before in this context. Imposing our own conversion, with a factor of 4.184, is also likely to play hob with precision. Saying that a given quanitity of food has about 200 calories (as sources are likely to), is not the range of values as "about 836.8 kJ". Septentrionalis PMAnderson 05:52, 19 January 2011 (UTC)

I believe that UK food labels now give kJ. I thus have to wonder if kilocalories should only be listed first in US-related articles and that UK-related articles or general articles should give kilojoules first. Jc3s5h (talk) 11:29, 19 January 2011 (UTC)

Calorie (food)#Nutrition labels. In the U.S., kilocalories, kilojoules, and even bothering to capitalize "Calorie" is a discussion found only in physics books, and non-scientific people will never encounter it. Thus our dieting article says "Low-calorie diets usually produce an energy deficit of 500–1,000 calories per day", which is what you call "kilocalories", but any dieting book or nutritional information I know of calls simply "calories". Uncapitalized. I'm not saying that's ideal; I'm saying that's the problem. Art LaPella (talk) 15:07, 19 January 2011 (UTC)

Countries where kilojoules are the default unit outnumber those where the (kilo)calorie - however it is spelled - is used by about 100 to 1. There is no way that preferring (kilo)calories over kilojoules by default makes any sense at all. Roger (talk) 19:00, 19 January 2011 (UTC)

Dodger67, what is your source for this information? Also, this is an English-language encyclopedia, so it would be interesting to try to estimate how many readers of the English Wikipedia would be more comfortable with kilojoules vs. Calories. This isn't easy to estimate because some people who live where English isn't the first language might feel the English Wikipedia is better for some subjects, or just prefer to read about some topics in English. In any case, the number of people is more relevant than the number of countries. Jc3s5h (talk) 19:19, 19 January 2011 (UTC)

If the source gives calories, it may make sense to use that as primary & convert to kilojoules, but I'm opposed to a blanket statement that (kilo)calories are to be the default preference. Default to the SI unit. Let's not assume that readers would be more likely to understand the ill-defined calorie. In Australia, for example, food is labelled in kilojoules. The conversion factor should rarely if ever be a problem since any decent conversion should be rounded apropriately. 200 4°C calories = 840.8 kJ whilst 200 20°C calories = 836.4 kJ but round them off & you get 840 kJ either way. I wouldn't bother with the "calorie"/"Calorie" thing. No everyone knows this absurd convention so we'd have to explain it anyway, so, why not just state that in this article we mean this by this? Though, if we've got a conversion, we might not really need any clarification. P.S. they use joules and electron-volts in physics. JIMp _talk·cont 21:59, 19 January 2011 (UTC)

I opppose the dogmatism that says we should default to SI - whether our readers will understand us or not. Septentrionalis PMAnderson 15:00, 28 January 2011 (UTC)

In the UK the style used to be Calories (for kcal) and began to shift towards kcal around ten years ago, the kcal now being the most used. The reason was that the general public, and indeed manufacturers/media, were having difficulty differentiating between 100 calories and 100 Calories - often saying things such as "One slice of bread is 140 calories" as opposed to the correct version which would be "one slice of bread is 140 Calories". Chaosdruid (talk) 17:09, 28 January 2011 (UTC) NB Having just examined three food items, bread, biscuits and fish fingers I can confirm that they are all given as "XXX kJ/XXX kcal".

• I agree 100% with PMAnderson, above. Just because kilocalories are—as Sarah Palin might say—“All scientificy” is no reason to push for the exclusive use of kilocalories in place of plain-speak. It should depend on the article.

  As a medical researcher specializing in metabolism and weight loss, I know a bit about this subject and can relate to writing white papers for a general-interest readership as well as for an expert readership; they are two different things. If one were writing an article on Metabolic energy requirement, which we don’t have although we do have Energy balance (biology), then writing “kilocalories” might be appropriate. But even still, the principal always applies that Wikipedia is directed to a general-interest readership and is not a scientific journal. So “kilocalories” in a more scientifically toned article would, IMO, be properly introduced with a parenthetical like A daily expenditure of 2200 kilocalories (2200 dietary calories). Conversely, for an article that will receive a high proportion of a non-expert readership, such as Morbid obesity, the use of “dietary calorie” should, IMO, come first with the “kilocalorie” being the parenthetical.

  This all falls under the same principal that for an article like Obesity, we write 2200 dietary calories per day and not the 2200 kcal·day^–1 that some editors must think makes them seem like “They must be from the big city.” Wikipedia does best when it uses appropriate plain-speak for the most likely readership and doesn’t try to promote the adoption of way-cool ideas—even if it’s the SI. Greg L (talk) 01:08, 29 January 2011 (UTC)

  
  P.S. FWIW, editors need to make an extra effort to not think of Wikipedia as a single entity requiring perfect internal consistency amongst its 6,843,515 articles. There are oodles of inconsistencies where non-SI units of measure typically have symbols that do not follow the rule of the SI one twit. No one interested in nautical technology will ever run off thinking that U.S.S. Enterprise can travel 35 nanometers in an hour. “Yes, but the real world is fu**ing retarded” you might say. Fine; follow the real world—and it’s retarded too. The job of any good encyclopedia is to follow the real world and use the conventions used by modern, most-reliable sources so our readers are properly primed for their continuing studies elsewhere on the subject. When it comes to diet & exercise and metabolic energy requirements and whatnot, no one uses the gram-calorie. The phrase consume 2100 calories per day is perfectly clear to our novice and expert readers alike. Whether it is “calorie” with a capital “C” or lowercase doesn’t matter; just look to the sources cited in particular article and follow the convention used by the majority of the best ones. Greg L (talk) 02:58, 31 January 2011 (UTC)

My two cents

1. No-one actually uses the small calorie of ~4 J or the kilolarge calorie of ~4 MJ in the context of nutrition, making calorie (Cal) and kilocalorie (kcal) effectively perfect synonyms at ~4 kJ in that context. The choice between the two should be governed by WP:RETAIN. (In my experience “calorie” prevails as the full name especially in spoken language, whereas “kcal” prevails as the symbol, much like many if not most people still call the micrometre a “micron” even if they don't use a lone “µ” as the symbol any more.)
2. While in many countries the kilojoule is ‘officially’ the standard unit for food energy, people still ‘think’ in calories. For example, in the EU food labels give both kilojoules (because they have to) and kilocalories (because they are the unit the costumers will actually read), and I don't think I have seen any book or any magazine article about nutrition using kilojoules. But I've heard that the situation is different in Australia. So, whether kilojoules or kilocalories are used in a WP article, there should always be a conversion to the other unit. A. di M. (talk) 11:06, 1 February 2011 (UTC)

... and if there is a conversion to the other unit, it will be obvious whether it's a kilo-gramcalorie or a kilogram-calorie. JIMp _talk·cont 21:43, 15 February 2011 (UTC)

In the context of nutrition, then, the "calorie" or "dietary calorie" ("Cal") is the large one and the "kilocalorie" ("kcal") is a thousand small ones. I'd say pick one and use it throughout the article be it the unit you're converting to or from. There should be conversions to/from SI. Is the unit used outside of these contexts? If so, it should still be converted to SI. Make an attempt to follow the definition of "calorie" used in the sources. JIMp _talk·cont 22:05, 15 February 2011 (UTC)

After a breif search of a few hundred articles it seems that "calorie" is much more commonly used than "kilocalorie". The number of WP hits are listed below.

• 91 for kilocalories
• 51 for kilocalorie
• 3,049 for calories
• 3,049 for calorie

Almost all uses of "calorie" appear in the context of nutrition. In this context the word generally (if not always) either refers to the large calorie or is used as a synonym for "food energy". The use of capitalisation to distinguish the large calorie from the small is rarely seen.

None of this is surprising when you consider that the calorie is essentially obsolete outside of nutritional contexts and that the small calorie is not of an appropriate size in these contexts.

Thus on thousands of pages (it seems) "calorie" does not refer to a gram-calorie as specified here. The word is ambiguous but we're not helping by insisting on conventions which will never be followed out there in the encyclopædia. And it's not likely that they ever will be followed out there in the encyclopædia whilst they aren't followed out out there in the World.

In the end, though, the solution is simple. Provide a conversion to SI and it will be clear enough which calorie is meant. JIMp _talk·cont 22:35, 16 February 2011 (UTC)

I propose the following wording.

The use of the calorie is dependant on context. When used in scientific contexts such as chemistry the calorie refers to the gram calorie (~4.2 J, symbol cal, also known as the small calorie). However, in nutritional contexts the calorie refers to the kilogram calorie (1000 cal, ~4.2 kJ, symbol Cal, also known as the large calorie, food calorie or dietary calorie).
∘Metric prefixes may be applied to the gram-calorie to form units such as the kilocalorie (symbol kcal). Do not apply metric prefixes to the kilogram-calorie. In all contexts the kilocalorie refers to one thousand gram-calories (1000 cal, ~4.2 kJ).
∘Don't use the large calorie and the kilocalorie in the same article except when comparing the units themselves.
∘Conversions to SI units (joules, kilojoules, etc.) should be provided.

Note that I've removed the links to "gram calorie", "large calorie", "kilocalorie", etc. & removed any suggestion that they be made in articles. They're all redirects to the main article, which doesn't even have a section devoted to the distinction. Also conversions to SI are enough to distinguish the big from the small. I've also added that the large calorie and the kilocalorie shouldn't be used in the same article, that would just be confusing. JIMp _talk·cont 01:57, 17 February 2011 (UTC)

The calorie, depending on context, may refer to the gram calorie or the kilogram calorie.
  ∘ The gram calorie (symbol cal, also known as the small calorie) is approximately 4.2 J. Metric prefixes may be applied to the gram calorie to form units such as the kilocalorie.
  ∘ The kilogram calorie (symbol Cal, also known as the large calorie, food calorie or dietary calorie) is equal to one thousand gram calories (1000 cal, ~4.2 kJ). Do not apply metric prefixes to the kilogram calorie.
  ∘ The kilocalorie (symbol kcal) always refers to one thousand gram calories (1000 cal, ~4.2 kJ).
  ∘ In scientific or technological contexts (such as chemistry or nuclear energy) the calorie (cal) refers to the gram calorie.
  ∘ In nutritional contexts the calorie (Cal) refers to the kilogram calorie. The equilavent kilocalorie (kcal) may be used instead but not both in the same article.
  ∘ Conversions to SI units (joules, kilojoules, etc.) should be provided.

New version. JIMp _talk·cont 13:42, 17 February 2011 (UTC)

"Mid" or "mid-"

The page doesn't specify a preference for the use of a hyphen when the adjective "mid" is added in front of years or decades. From what I can tell, "mid-1990s" is preferable to "mid 1990s". Is that correct? (See also Hyphen#Prefixes and suffixes, which isn't very helpful.)   Will Beback  talk  21:27, 19 January 2011 (UTC)

Some dictionaries mention this sense of "mid" as a word and not a prefix, but some don't. A Google Books search prefers "mid-1990s" by about 10 to 1, and Google Scholar by about 3 to 1. So I would say "mid-1990s" is preferable. Art LaPella (talk) 00:54, 20 January 2011 (UTC)

Confederate banknotes

I just passed this subject article from idle interest, and am unfamiliar with the fractional templates, so if there's a consensus, perhaps someone here might be interested in fixing (or at least harmonising) the tenths and halves at Confederate States of America dollar#Banknotes —— Shakescene (talk) 21:24, 21 January 2011 (UTC)

On examining images of some fractional Confederate dollars, I saw that they were denominated in cents, rather than fractions of a dollar, so I just changed them to "10-cent" and "50-cent". The question is now moot.

Resolved

—— Shakescene (talk) 00:17, 23 January 2011 (UTC)

MOS vs. MOSNUM on centuries

At the moment, MOSNUM says "Centuries are given in figures or words using adjectival hyphenation where appropriate: the 5th century BCE; nineteenth-century painting." The same section in MOS (WP:MOS#Numbers as figures or words) says "Show centuries in figures". Who wins? - Dank (push to talk) 02:21, 22 January 2011 (UTC)

Chaos wins. Art LaPella (talk) 04:40, 22 January 2011 (UTC)

Better chaos than Newspeak. MOSNUM (uncharacteristically) describes what English actually does; MOS prescribes what some editor thinks we all ought to do; while endemic, this is contrary to policy. I prefer the former, and intend to ignore the latter; but it would be better to copy the longer and more accurate text to WP:MOS. Septentrionalis PMAnderson 14:58, 28 January 2011 (UTC)

It needs to be made consistent. Consensus needs to be generated for either version, and used in both. Tony (talk) 15:01, 28 January 2011 (UTC)

Since they have been inconsistent for a considerable time (and nobody has noticed), the necessity is unclear; but since one is preferable to the other, I would agree to change MOS. Septentrionalis PMAnderson 16:03, 28 January 2011 (UTC)

And I have; comments? Septentrionalis PMAnderson 02:33, 29 January 2011 (UTC)

And been reverted on autopilot. Is there any reason at all to suppose anybody actually supports the MOS text? Let inconsistency reign, as it has for months. Nobody noticed before; nobody cares what these texts say now. Septentrionalis PMAnderson 03:50, 29 January 2011 (UTC)

Thank you for pointing out the inconsistency between the pages. It needs to be discussed, widely, since ordinals and centuries are widely used. Alas, some articles have them written out. I think the best place to work out what should be done is at WT:MOS. Tony (talk) 05:07, 29 January 2011 (UTC)

The inconsistency isn't just between pages. After the so-called "autopilot" edit linked above, there is now inconsistency within the same sentence. "*Show centuries in figures: (the 5th century BCE; nineteenth-century painting)." In other words, it doesn't matter whether Wikipedians "Show centuries in figures" or write "nineteenth-century painting" instead; do we play this game just so Manual of Style regulars will know whose side we're on? If we can't keep MOS from contradicting MOSNUM, can't we at least be consistent within the same sentence? Art LaPella (talk) 05:44, 29 January 2011 (UTC)

• Instead of wrangling about the above procedural issues, why not get to heart of the matter? Should the MOS and MOSNUM both state that 'only numbers are to be used for century' or should they both state that either 'numbers or words can be used for century'. The latter results in what I have found thoughout WP articles: in the same article (or even the same paragraph or sentence) a back and forth mixture of century numbers and and century words. One editor even stated he was deliberately writing articles in this way so his readers would not get bored just seeing century numbers and not words. So what should be written here? Hmains (talk) 06:38, 29 January 2011 (UTC)

  THe solution to inconsistency within an article is simple. Often it doesn't matterl if the two centurys are at opposite ends, it would take a search engine to notice. If it jars, apply WP:CONSISTENCY, using whichever is most suitable to the places where it occurs; since 19 is a small number; nineteenth is usually best by our general principles on figures and words. Septentrionalis PMAnderson 17:56, 30 January 2011 (UTC)

  • Because for me at least, the "procedural" issue is the heart of the matter. Even for a partisan of the "sixth sentry but 6th century" rule, the heart of the matter should be what happens in the rest of Wikipedia, not here, and Wikipedians can't possibly obey "Show centuries in figures: ... nineteenth-century". One editor doesn't want his readers to get bored? Sounds like a good reason to me, but then I don't write the Manual of Style guidelines other than "procedural" issues. If you don't like "nineteenth-century", please change the whole sentence or none of it, and change MOSNUM to match. There's no reason to change part of the rule but not all of it. Art LaPella (talk) 17:06, 29 January 2011 (UTC)
• No. We certainly should not prescribe that centuries should always be in figures; for one simple reason: It's not English; 19th century is enormously less common than nineteenth century, even in the twenty-first century. I am extremely tired of MOS provisions which cause bots to go out and make Wikipedia look stupid. Septentrionalis PMAnderson 17:41, 29 January 2011 (UTC)
  • "It's not English"? "19th century" makes WP look "stupid"? Tony (talk) 14:06, 30 January 2011 (UTC)
    • Yes, precisely. Never to use "nineteenth century", even where it is idiomatic, is as odd-looking as George Bernard Shaw's omission of apostrophes: dont, cant, wont. To forbid the normal and prevalent usage with no rationale is also another of the Manual of Style's innumerable pieces of Original Research. Septentrionalis PMAnderson 17:35, 30 January 2011 (UTC)
    • I see no reason not to tolerate 19th century as occasional usage; it does appear to be used - now - about 15% of the time, and it will occasionally be actually preferable (infoboxes, six 19th-century paintings and so on), but the MOSNUM text does that. Septentrionalis PMAnderson 17:46, 30 January 2011 (UTC)

I think MoS:Numbers and MOSNUM should be looked at like "==Numbers==" and "{{Main|MOSNUM}}". In other words MoS should be a short section summary of the main points in MOSNUM.

My point is that MoS is a general editing guide which has general rules on style etc., and MOSNUM is the place where that style is further defined. If we consider how each separate entity (such as projects) treat MoS, having their own guidelines specific to their project needs, MOSNUM is simply another "project" dealing with numbers as opposed to MilHist, Countries, BLP or Television which deal with those specific topics and produce their own MoS'.

I, for one, have a great deal of interest in the fact that MoS and MOSNUM say two different things, and in getting the matter clarified, as I have to copy-edit articles. While there is the obvious matter of consistency, the problem delves into the realms of English English v. American English and all the extra rules that brings, "Whichever was first the rest shall be the same" etc.

If one considers the styles in which people edit it seems to me that most editors will write "5th–7th century" and "17th–19th century" rather than "Fifth to seventh century" or "seventeenth to nineteenth century" as it is easier to write and to read. When quoting single periods it seems that the general trend is to mix and match, and certainly the articles I have edited seem to follow those trends.

In conclusion I think it would be wrong to force one or the other and that MoS should take the lead from MOSNUM. There could be an argument made that consensus should be given to a larger group than simply those involved in MOSNUM, yet the case is that the Jigsaw method works—as in the Calorie/kcalorie/kj where only a few people are really interested in discussing it: Similarly the majority of editors are not going to go into each corner of Wikiworld to enter discussions on every point of style for each of the projects, MoS groups or discussions as the time taken to catch up reading 50 conversations a day would stop them from editing and Wiki would be left to the hands of the vandals, unsourced-text dumpers and OR-mongers.

I also think that using C15th is best left out of the picture :¬) Chaosdruid (talk) 19:37, 30 January 2011 (UTC)

• Really? There's this much debate about taking a sentence that contradicts itself and at least fixing it so that the sentence has internal logical consistency? WP:MOS now says "Show centuries in figures: (the 5th century BCE; nineteenth-century painting)" Now, according to my copy of The New Oxford Dictionary, a figure is "a numerical symbol, esp. any of the ten in Arabic notation : the figure 7"—at least, that's the only of the five meanings remotely applicable to a century. "Nineteenth" is a word, not a figure, by the dictionary definition of figure. Therefore, as currently written, this guideline tells editors in clear language to do one thing... and then, by example, directs them to do just the opposite. When someone tries to correct this obvious, glaring, and incontrovertible logical error, the change is reverted because "that's the way it's always been". While it's fine to debate whether or not the rule should be changed, the fact stands that the example given flatly contradicts the explicit instruction of the text. There's no reason not to fix that right now by altering or removing the faulty example... yet doing so seems to be anathema here. Instead, there's debate about changing the rule to fit the example. Am I the only one that's questioning the sanity of this whole process? Maybe we should fix the error, then debate changing the rule? // ⌘macwhiz (talk) 23:15, 30 January 2011 (UTC)

• (Like Sioux indians sitting around the campfire on ‘Dances With Wolves’): User:⌘macwhiz speaks strong words. Greg L (talk) 00:14, 31 January 2011 (UTC)

• IMO, numerals are much easier to read than spelled-out words when acting as numerators for centuries. A few examples above show nicely how this is especially the case in expressing ranges. Macwhiz, I reverted the change because the whole phrase needs to gain consensus. I know it was a glaring illogicality, and that it should have been addressed long ago. Tony (talk) 02:24, 31 January 2011 (UTC)

• A good and proper move. Greg L (talk) 02:43, 31 January 2011 (UTC)
• In other words,. if that were so, this guideline (and MOS) is based on Tony's opinion. No wonder he's always concerned about its status.;} We have a term for that: {{essay}}. Septentrionalis PMAnderson 02:52, 31 January 2011 (UTC)

• Maybe I’m confused then, PMA. Please provide the disputed diff. Greg L (talk) 03:06, 31 January 2011 (UTC)
• So am I. Tony doesn't seem to have actually reverted anything. Septentrionalis PMAnderson 03:17, 31 January 2011 (UTC)

I have restored from this this revert on MOS because the edit summary This is how MOS has read for many years, prior to any of us being involved. Discuss changes, don't push) is a falsehood; Art LaPella is also correct that the text is incoherent. That section, as it stands, is no older than its last serious rewriting, two or three years ago, in which both Tony and I were involved. I presume, since both of us missed it, the present garble is younger still. (The revert-warrior was also editing then; whether he was involved with MOS I do not recall.) Septentrionalis PMAnderson 03:13, 31 January 2011 (UTC)

• I tried to get this actually discussed instead of raids on the text. Alas, no success. The examples do not match guideline text: fix the examples is the obvious conclusion. Instead, the raids change the guideline. But I wanted to first decide on the guideline; the examples will follow in turn. However, instead of discussion, we get insults. I take that to mean that there is no foundation to the arguments. Insults cannot be taken seriously, nor the insulter. Do better. Hmains (talk) 03:18, 31 January 2011 (UTC)
  • We have discussed it, here and briefly on WT:MOS. You are the only person who actually supports this piece of Original research, contrary to English usage. Your reasoning is "in the same article (or even the same paragraph or sentence) a back and forth mixture of century numbers and and century words." That is a fallacy; by the same reasoning, we could justify a rule always use favour because ENGVAR could lead to in-article inconsistency. So it could; the answer is not to impose a straitjacket, but to be consistent. Septentrionalis PMAnderson 03:31, 31 January 2011 (UTC)
  • "How MOS has read for many years"? Hmains might make some sense if that were true, but it isn't. Which year? 2010? That says "19th" not "nineteenth". 2009? "19th".2008? "nineteenth" only in a special case, and not self-contradictory at all. 2007? "19th". 2006? No century rule at all. Art LaPella (talk) 03:57, 31 January 2011 (UTC)
• Very well. I see that PMA’s edit addresses macwhiz’s spot-on observation that the example given flatly contradicts the explicit instruction of the text. And I don’t have a problem with what’s there now; neither figures nor words confuses me one twit. Sometime, depending on adjacent text, one form might be preferred over the other. Oh, Tony’s got a Ph.D. and makes a living as a professional technical writer so he’s not exactly a lightweight around here. As one of ‘them edumacated types,’ I find that he is typically *keen* on the general rule and is also readily able to detect the exceptions to the rule that are usually only appreciated by the *finer* sources on the subject. That doesn’t mean he floats when he meditates, but it does mean one would be wise to double-check the books on the subject, PMA, before suggesting he’s pulling any of this stuff out of his butt. Greg L (talk) 03:26, 31 January 2011 (UTC)
  • I didn't say he had; but his post claims the right to do so. I prefer evidence. Septentrionalis PMAnderson 03:31, 31 January 2011 (UTC)
    • Oh, pardon me. I had loosely translated {{essay}} as “pulling stuff out of one’s butt.” Too much of a leap? Greg L (talk) 03:38, 31 January 2011 (UTC)
      • Yes. essay means "somebody's opinion" and no more. Some essays are very worth reading, but this is a guideline.
      • And if we're going to rely on opinion, I find the numerals too visible, indeed glaring; they make text harder to read. Why forbid 5 pigs and demand 5th century? Septentrionalis PMAnderson 03:41, 31 January 2011 (UTC)
        • I tend to agree with you (and macwhiz, who conveniently quoted the Chicago Manual of Style, below). But I also know that there are those who will speak about “tables” and other areas where “space is at a premium” (yadda yadda yadda ad nauseam). This issue isn’t important enough, IMO, to put up with all the Turkish butt-stabbings and amputations of body parts by ArbCom. The current version you reverted to is no longer internally inconsistent and is good enough. There are simply too many contributors on Wikipedia with different ideas and the faux pax is a grey area with nuances. Greg L (talk) 03:47, 31 January 2011 (UTC)
  • (EC) Speaking of which, I just pulled out Chicago Manual of Style, 16th Edition to see what guidance it might give. Section 9.33 is unequivocal on the subject: "Particular centuries are spelled out and lowercased." While CMoS isn't controlling on the MoS or Wikipedia, it's extremely compelling evidence for "this is the proper way to do it in American English". I'm also glad to see that we seem to be building consensus that sentences in the Manual of Style should not be schizophrenic, whether they be dogma or heresy. I was worried that we might need to call an RfC on that one for a while... // ⌘macwhiz (talk) 03:35, 31 January 2011 (UTC)
    • Would that be Chicago Manual of Style, 16th Edition or Sixteenth Edition? :) TechnoSymbiosis (talk) 04:24, 31 January 2011 (UTC)
      • Since you mention it... 16th Edition. Section 9.6: "The general rule applies to ordinal as well as cardinal numbers." The "general rule" for Wikipedia being CMoS's "alternative rule" 9.3, that zero through nine are spelled out, and all other numbers use numerals. (Chicago prefers spelling out zero through 100. I find that a bit bizarre, personally.) However, that's under the heading General Principles; section 9.33 overrides 9.6 when it comes to centuries. // ⌘macwhiz (talk) 04:44, 31 January 2011 (UTC)
        • • Presumably like Volume numbers (9.28): "usually...in Arabic numerals". Septentrionalis PMAnderson 07:17, 31 January 2011 (UTC)
• Why 5th century but not 5 pigs? Well, 5 pigs and 12 horses, apparently, is what is mandated to avoid clash between numbers and words, although I've never thought much of that. And as Greg points out, in tables (and captions, where we see a lot of "19th-century this and that"), the numerals are much neater and easier to identify. They do not seem to stick out in the longer unit that arises from the ordinal suffix plus "century", and are much used in this context. Tony (talk) 08:15, 31 January 2011 (UTC)

  Both forms are useful. Which to use in any given context is editorial judgment. Don't let's jog editorial elbows. Septentrionalis PMAnderson 19:51, 31 January 2011 (UTC)

• EXECUTIVE SUMMARY: I agree more with Tony here. I find that numeric centuries look more natural. And it might be that the reason for this ‘naturalness’ is that numeric centuries is a far more common writing style in the real world (by a factor of nearly 30). If we were to change the guideline, I would propose that we say either is acceptable, but to suggest a preference for editors to simply follow the writing style used by a majority of most-reliable sources cited in a given article.

  ---

  If I had only been exposed to examples of real-world practices and had no knowledge of how various manuals of style addressed this issue, I might write 19th-century engine refrigerates without CFCs, as did this New Scientist article. New Scientist, as I recall, is written using British English if that matters at all. I find that expressions like Galileo Galilei was a famous seventeenth-century inventor is a tediously long. The tediousness of the ‘teen’ centuries might underlie why we have Category:17th-century astronomers. At least, numeric centuries work in titles (as it did in New Scientist). Clearly though, numbered expressions like …theory of the 16th-century astronomer Nicolaus Copernicus… are used also in body text by seemingly professional looking publications like this article from The New York Observer.

  I don’t know why, exactly, but I think I prefer the numeric centuries myself—as does Tony. Perhaps our eyes are simply more used to that form: Witness this Google search on "Sixteenth-century astronomer" (in quotes). That one returns 15,300 results. But "16th-century astronomer" returns 445,000 reults; 29 times more frequent.

  If I had to throw out a proposal to change the guideline, it would be to follow the practices of the majority of most-reliable sources used in a given article. Practices might vary depending upon whether it is an archeology-based article, or an astronomy-based one, etc. This approach would best leverage the mission of any good encyclopedia: to properly prepare its readership for their continuing studies on the subject—writing conventions (writing style) and all. The outcome of such a guideline change would be that most articles would probably use numeric centuries. Greg L (talk) 20:25, 31 January 2011 (UTC)

Possibly subject matter. But also the writing style of the web; Google Books is two to one the other way. If our editors want to avoid typing seven characters, why force them? but that way lies l33tspeak.Septentrionalis PMAnderson 23:02, 31 January 2011 (UTC)

• I see that you provided a link wherein you unintentionally did a Googlebook search that was a lowercase “L” as part of the “16”, as in l6th-century. When done properly without the lowercase L, so the number 16 is an actual 16, one gets the results you cited; a 2:1 preference for "sixteenth-century astronomer", which returns 426 results, vs. "16th-century astronomer", which returns 199 results.

  I agree with your point; that venue (Internet vs. actual books) account for a humungous difference in the outcome, with a 2:1 preference for “nineteenth-century” in books and a 29:1 preference for “19th-century” on the Web. I also note that this string "in the 19th century" yields 749,000 results whereas "in the nineteenth century" returns 7,820,000 results.

  This may be as simple as book writers tend to be professionals who have actual manuals of style, vs. the Internet, where everyone’s an expert simply because they have an X-chromosome, a pulse, and an opinion. It would be a phenomenon similar to “gigawatts” (before computers got to giga-anything); it was properly jiga (as in “gigantic”). But because the the prefix “giga-” worked its way down the food chain into consumer-grade products and all the way into the Wal‑Mart crowd buying $549 Dell computers, it became—due to simple cluelessness—“giga as in biga which is gooder”.

  I must be from the Wal‑Mart set, since I prefer the look of numeric centuries notwithstanding the fact that grammar-school teachers would whack my knuckles with their rulers. Maybe it’s funner that way. Greg L (talk) 00:29, 1 February 2011 (UTC)

• (edit conflict) Google hit counts for multi-word searches are only estimates, and when they are larger than a few thousands they are often off by several orders of magnitude.[1] A. di M. (talk) 01:00, 1 February 2011 (UTC)

(edit conflict) My two cents: the current guideline (‘Centuries are given in figures or words&nsbp;...’) is no more informative than having no guideline at all. (Of course they are given in figures or words; how else could they be given? Egyptian hieroglyphs? IPA transcriptions of their names?) So I'd propose just removing it. (The consequence would be that the generic guideline would also apply for centuries: prefer words up to the ninth century, don't use words for some centuries and figures for other ones in the same context, etc.) A. di M. (talk) 01:00, 1 February 2011 (UTC)

Reason: to stop editors fighting about it ... so they know. But Greg L's research above strongly suggests that centuries should be qualified by numerals, not words. I see numerals all over the place in WP articles. Tony (talk) 01:11, 1 February 2011 (UTC)

FWIW,

• in the British National Corpus, excluding the "spoken" and "fiction" sections, "first century" etc. are more than 10 times as common as "1st century" for centuries up to the 9th, between 2.5 and 6.4 times as common for centuries from the 10th to the 20th, and about 30% less common for the 21st century;
• in the Corpus of Contemporary American English, excluding the "spoken" and "fiction" sections and only including the "2005–2010" part, "first century" is about 16 times more common than "1st century", then this ratio gradually gets smaller for later centuries, being between 1.2 and 1.9 for all centuries from the 10th to the 19th, it is about 1.1 for the 20th century and about 0.5 for the 21st century. A. di M. (talk) 02:20, 1 February 2011 (UTC)

• Nice research. Thanks. Is anyone proposing that the current guideline be changed? It seems both “fifth century” and “19th century” are here to stay. Should we adopt the recommended practice as shown in The Chicago Manual of Style because spelled-out centuries are clearly preferred by choosey writers? Or, in acknowledgement of the ubiquity of the numeric centuries when it comes to on-line internet use, shouldn’t we allow both? Greg L (talk) 03:46, 1 February 2011 (UTC)

  I'd be happy with not having any rule specifically for centuries at all (so the general rule of preferring words for smallish numbers and figures for largish ones, but not at the cost of using words for some centuries and numbers for others in the same context, would apply); but if we really must have one ‘to stop editors fighting about it’, it should be allowing both. A. di M. (talk) 09:58, 1 February 2011 (UTC)

User:Noetica has been watching this thread and emailed me the following message, for posting here:

Here is one of the covert and dishonest edits accepting words for centuries, leading to the present disarray: [2]

Here is the archived discussion (at that time 12 months old) linked to in the summary for that edit, and ostensibly justifying it: [3]

At the time of that edit, you [Noetica is referring to me] challenged it: [4]

The perpetrator made a good solid answer to you. But while he did mention a need for consistency, nowhere in his enumerated points or in the many guides he cites is there any rebuttal of our salient point: every publication sets ONE standard for naming centuries, not two. This is what I urged in the 2009 discussion that is linked above. Why the bejesus should Wikipedia be the sole exception, permitting both? In fact, WP practice where it matters is solidly in favour of figures, not words. For example: List_of_centuries, Category:Centuries, Mid-nineteenth_century_Spain.

(That last one redirects to Mid-19th-century Spain. So it damn well should, if we are to have any evenness at all in our style.)

Many publications choose one way of spelling nineteenth century (experience casts some doubt on whether they get it). Most publications also require one specific way of spelling flavour; we expressly refuse to have one, in one of the few actually widely accepted MOS provisions. We are a new thing, a massive international collaboration; we are dealing with a question in which English usage varies, and preference may reasonably depend on context. We should not exclude the greatly more common usage without clear argument of benefit to the encyclopedia, and "evenness of style" (when we have not attained it, or reject it, in far more obvious matters) is buncombe. Septentrionalis PMAnderson 22:45, 1 February 2011 (UTC)

I see that Andy Walsh, at the first link-target, asked who wants "twenty-first-century music". A very reasonable question. Tony (talk) 12:08, 1 February 2011 (UTC)

Good point, but that's no good reason to forbid spelling out the numbers (e.g.) in an article about ancient history which never mentions any century later than the seventh. (If we had to have an iron-clad rule, I'd propose that articles mentioning at least one century equal to or later than the 21st must use figures, articles which never mention any century later than the 20th and mention at least one century equal to or earlier than the 9th must use words, and articles which only mention centuries between the 10th and the 20th must use WP:RETAIN.) --A. di M. (talk) 14:02, 1 February 2011 (UTC)

I am still convinced that the two word rule is imperative also.

I think that ranges should be numeric: 15th - 17th centuries. They are more than four words to express the range and, as we do not write fifteenth November eighteen-hundred, this is more in keeping with the guidelines on Other date ranges

MoS also states "As a general rule, in the body of an article, single-digit whole numbers from zero to nine are spelled out in words; numbers greater than nine are commonly rendered in numerals, or in words if they are expressed in one or two words" Numbers as figures or words

The whole thing is a massive confusing set of "rules" that state "do it either way".

A second (and I hear someone groaning at the back) point is that eighteenth century is 6 characters longer than it needs to be. Consider if we were to remove an average of one entry per page in the encyclopaedia as a whole - that would be a lot of space saved - wouldn't it? Well we would have to do that to 4,820 articles just to free up the same space as this discussion lol - not a convincing argument for that then :¬) Chaosdruid (talk) 16:08, 2 February 2011 (UTC)

No space is saved if any instance of "eighteenth" is replaced with "18th", as the old revisions of the article are stored as well. --A. di M. (talk) 08:54, 3 February 2011 (UTC)

DOH! I completely forgot about that lol - thanks for pointing it out :¬) But then, after thinking about it, wouldn't each subsequent revision have 6 less, so after 20 changes the saving would be 20x6 ? Chaosdruid (talk) 16:37, 3 February 2011 (UTC)

Even in a 6-kilobyte article (a pretty short one) you'd need one thousand more edits before having a net saving in space. :-) Also, I believe that the Wikipedia servers have tens times as much disk space available than is currently used (though I can't find the relevant statistics right now), making such arguments pretty much irrelevant. --A. di M. (talk) 17:30, 3 February 2011 (UTC)

Currency - always have country in front of dollar sign

Shouldn't the country sign always be in front of the dollar sign (US$, S$, A$ etc.). If it's just the first instance, people starting to read at a later section won't know for sure what currency they are reading about. It will also help highlight any mistakes. Certainly if you look at the style guides of serious financial publications - FT, Reuters, Economist etc., I'm pretty sure that would be the case, and surely wiki should be trying to keep to the same high standards. Mattun0211 (talk) 07:48, 2 February 2011 (UTC)

MOS:CURRENCY says, "Use the full abbreviation of a currency on its first appearance (e.g. AU$52); subsequent occurrences can use just the symbol of the currency (e.g. $88), unless this would be unclear." If the whole article only uses one currency then there would be no problems. It's a judgement call as to whether it would be unclear or not. McLerristarr | Mclay1 08:28, 2 February 2011 (UTC)

IMHO it's very much dependent on the context, for example in a discussion of the budget of a city or state/province it would be irrational to assume that anything other than their own currency is meant. For example the budget of the city of Sydney, Australia, would be discussed in terms of AU$, while the Ohio Department of Education's expenditure would be expressed in US$. It shouldn't be necessary to constantly specify which dollar is meant. International trade or comparisons between different countries are often expressed in terms of US$, although in some contexts other currencies such as Euros would be used. Roger (talk) 08:57, 2 February 2011 (UTC)

The issue is whether the reader will also assume the same. Someone might well have no idea that the Australian currency is also called a dollar, or even that there's any other currency than the US one using the dollar sign, so on reading "$10,000" in an article about Australia they might well think that US dollars are meant. --A. di M. (talk) 12:09, 2 February 2011 (UTC)

Per MOS:CURRENCY, the first instance must specify which dollar (AU$, US$, etc.) so that solves the problem. I don't agree that the prefix is needed at every instance in an article - unless the article is specifically using multiple currencies. Roger (talk) 13:24, 2 February 2011 (UTC)

I think you're missing the point I was making Dodger. While we all like to assume that the reader is reading every word of our elegant prose ; in reality, most people skim read. As such, there is a fair chance they will miss the first instance, and will even make a wrong assumption or be confused when they come across the later instances - I know I've been there myself on Wikipedia. This is why respectable financial publications such as the Economist, FT and Reuters have a rule of always making this clear - even when it may seem obvious, as otherwise there would be a grey area of when to and when not to apply the rule - and what's obvious to one person may not be obvious to someone from another country or someone not familiar with the subject. This is not done for anacronyms as it would be too cumbersome, but asking someone to put an A in front of the $ sign, for instance, is hardly too taxing. Wiki style guidelines are generally spot on, but this one stands out like a sore thumb.Mattun0211 (talk) 01:50, 3 February 2011 (UTC)

• I've just criticised an article on Iran for using "US" before $, first time. I was under the impression the style guides said $ and ₤ needed no national tag, unless the currency of the country in question in the topic also uses "dollars". Even then, it's irritating to have A$ (or worse, AUD) repeatedly through an Australia-related article if there is no doubt (after the first occurrence) that all of the currency references are to Australian dollars. Tony (talk) 07:18, 3 February 2011 (UTC)

  I wouldn't assume any reader to know off the top of their head whether or not the Iranian currency is called a dollar, so I guess the first occurrence does need "US". (I'd also possibly use it for the first occurrence of it in any section directly linked to from another article or redirect, and for the first occurrence of it in each section if there's no occurrence of it in the lead–as I expect the reader, after reading the lead, might use the TOC to skip to another section.) --A. di M. (talk) 08:52, 3 February 2011 (UTC)

I think you're missing the point Tony. Reuters, The Economist and the FT, and I would suspect pretty much all financial publications with an international audience always specify the country before the dollar. As I've said twice, this is partly because people tend to skim read and you can't second guess where they're going to start skim reading from. It isn't annoying - if you read serious international financial publications you'll be reading this all the time (AUD isn't annoying by the way - it's just wrong from a style point of view). I suspect the current style came from people who were used to more parochial publications, probably in the US. Wiki is clearly a medium that is very much international. While there are times this may seem and even be relatively unnecessary, the rules that would govern that would be complex - one reason why the serious financial heavyweights don't try it. If it's good enough for the leading financial publications then we really should be following them rather than pretending we know best and trying to reinvent the wheel. That's my final word on the subject. You can lead a camel to water ... Mattun0211 (talk) 18:48, 4 February 2011 (UTC)

• Wikipedia is unlike Reuters or The Economist because we have articles that often make the type of current exceedingly obvious because of the nature of a given article. (From “Brooklyn Bridge”): The bridge cost $15.5 million to build and approximately 27 people died during its construction. It’s perfectly clear what the type of dollar is without the use of the prefix clarifier. That is something that is impossible with a Reuters article distributed across the globe, which tells of the recent increase in the price of a barrel of oil to a new record level. They clearly must specify what type of dollar is used for an international audience to understand a generic news bit. But, as can be seen from the Brooklyn Bridge example, Wikipedia often suffers no such ambiguity. If it isn’t clear from the topic and/or context, then clarify with the prefix. Otherwise, keep it *clean* (without) because most people read local (city, state, national-level) publications in daily life where no clarifier is used and the use of it here is unnatural and tends to distract by drawing the reader’s attention to the writing style rather than the message point. Greg L (talk) 05:53, 6 February 2011 (UTC)
  • Reuters could just as easily have a story where it was as obvious what the currency was as the Brooklyn Bridge example, but that's not the point. You say, "That is something that is impossible with a Reuters article distributed across the globe..." - but Wikipedia is distributed across the globe as well, is it not!? (And in the example you gave of a barrel of oil - oil is always priced in US$.) I agree, not having the country prefix is fine for publications that don't transcend national boundaries, but Wiki does - in spades. Do people find the country prefix unnatural when reading The Economist? I doubt it. They wouldn't even notice it. That standard, surely, must be what wiki is aiming for, rather than the Oklahoma Gazette. I live in the Asia-Pacific region where this problem is probably more acute as you have the Singapore Dollar, Hong Kong Dollar and Aussie Dollar, so there's probably more room for confusion than elsewhere. But I know I 've read wiki articles where I've been unsure what currency is being specified and recently a Chinese colleague was reading a wiki article and asked me what type of dollar was being referred to. Surely Wiki needs a proper style guide, not one that talks about the fact that is "often" exceedingly obvious, or asks writers to make judgement calls. There should be a style guide that is unambiguous and caters for the global readership whether they're in Singapore, London or West Virginia. Mattun0211 (talk) 02:34, 7 February 2011 (UTC)
    • I guess your Chinese friend will be left in the dust if he can’t figure out what The Brooklyn Bridge cost $15.5 million to build… means. We write MOS and MOSNUM rules for two or three standard deviations on each side of the bell curve. If someone has galactic-grade cluelessness and is outside of the 99.5th percentile, they are at risk of being left behind so we can have a writing style that is clear, crisp, and doesn’t draw undo attention to itself for the vast majority of our English-speaking readership. (Yeah, I know, “En.Wikipedia is *special* and is read by yak herders in Mongolia.”) I read about someone who adopted some Hmong refugees. Found them out crapping in the garden one evening. You know, even after just a day or two, they got the concept of “toilet” and we don’t have to link it. Greg L (talk) 03:32, 7 February 2011 (UTC)
      • It's plain annoying when The Economist insists on repeating ad-infinitum A$ or S$ when it's perfectly clear which dollar is intended. Tony (talk) 03:34, 7 February 2011 (UTC)
        • The only articles in which I would mention dollars without specifying which kind of dollars I mean (at least in the first occurrence of it in the lead and possibly in each section) are those using US dollars only and dealing with no other country than the US. Oh, wait: that's what MOSNUM already suggests. A. di M. (talk) 11:43, 7 February 2011 (UTC)
        • Agreed. If the writing style draws attention to itself to the point it’s annoying, it’s the wrong writing style—even if a yak herder for whom English is a second language might find it helpful. MOS:Currency seems to have had some rational thought behind what’s there and I don’t have a problem with it either. Greg L (talk) 03:39, 7 February 2011 (UTC)
          • Hi Tony. The vast majority of people wouldn't notice it. The reason (fourth time lucky!) is to take account of the fact that most people skim read, so you can't assume they will catch the first instance. I think you misread my post Greg - I didn't say my Chinese friend was reading an article about Brooklyn Bridge! To be honest, I'm beginning to tire of wikipedia due to exchanges like this. As we're talking about the Economist, perhaps you might find these articles [5] [6] interesting.

"It is the third worry—that Wikipedia has become ossified and bureaucratic, discouraging new users from contributing—that is the greatest cause for concern. In recent years its most active contributors have become obsessed with obscure questions of doctrine and have developed their own curious jargon to describe the editing process. The number of regular contributors to Wikipedia’s English edition peaked in March 2007 and has since declined by a third; the number of new contributors per month has fallen by half. Growth in the number of articles and edits has also levelled off. ... To ensure its long-term health, it needs to rediscover the flexibility of its early years."

"some evidence suggests that neophytes are being put off by Wikipedia’s clique of elite editors. One study by researchers at Xerox’s Palo Alto Research Centre looked at the number of times editorial changes were subsequently reversed. It found that roughly a quarter of the edits posted by occasional contributors were undone in late 2008, compared with less than 2% of those posted by the most active editors. And it noted that this gap had widened considerably over time." Mattun0211 (talk) 07:30, 7 February 2011 (UTC)

• I agree with Roger's first point above. As to Mattun's last point, it would be more interesting if Xerox and Mattun reported on "like" edits ... e.g., edits with refs. My guess is that regular editors much more often use refs than do occasional editors. Similarly, stats on vandalism in the two groups would be of interest -- again, I expect that overt vandalism reverts are a much greater ocurrence w/occasional editors. Nice numbers .... but without more, they tell me little.--Epeefleche (talk) 09:31, 7 February 2011 (UTC)

I really will call it a day on this one this time; All I'd say is that if you're going to be a serious international publication then you should have international standards. The key point I keep making about skim reading keeps being ignored. I didn't quite understand the edits with refs point, to be honest - I use references all the time when editing if that's what you mean.Mattun0211 (talk) 10:23, 7 February 2011 (UTC)

My point is that I expect that occasional editors are more likely to add text bereft of refs. In general -- that was not a comment on Mattun's editing at all. If that is the case, which I think likely, it is no surprise that text additions that lack refs are reverted with greater frequency than those that have refs. Which would explain some (if not all) of the difference in reversion rates. Best.--Epeefleche (talk) 18:00, 7 February 2011 (UTC)

What implies what, what?

Talk:Speaker of the New York State Assembly

Legislative terms use - (short dash) , life spans use – (long dash). Besides, 1820-21 is a one-year term which had sessions in two calendar years. 1820-1821 would be a two-year tenure. Besides, the dashes and numbers were messing up page references. Please check carefully what you are editing, the correct info should be preserved. Please avoid unnecessary edits. Kraxler (talk) 21:42, 29 January 2011 (UTC)

I guess I didn’t get that memo? ―cobaltcigs 01:48, 3 February 2011 (UTC)

Actually, that's a reasonable differentiation, between a session in both of two years (often only a small part of each) and a term which extends over a period of time. Since it will not convey the message to most readers, it should be amblified by saying session and term. Punctation is a tool, not an endurance test. Septentrionalis PMAnderson 19:22, 3 February 2011 (UTC)

This strikes me as an artificial and counter-intuitive distinction. An imprecise time-span is indicated by the year in which it begins and the year in which it ends, without regard to calendar month/day. That is, if an individual holds office from “January 3, 2004 to December 31, 2005” or from “August 8, 2004 to August 8, 2005” we should use the abbreviation “2004–2005” in either case. Anything more specific would and should require displaying the months for comparison. ―cobaltcigs 01:39, 4 February 2011 (UTC)

Euro plural

Why does the MOS recommend using the incorrect plurals "euros" and "cents" over the ECB's offically-defined plurals "euro" and "cent"? Stifle (talk) 15:57, 3 February 2011 (UTC)

http://ngrams.googlelabs.com/graph?content=50+euro%2C+50+euros&year_start=2002&year_end=2011&corpus=0&smoothing=3. A. di M. (talk) 16:17, 3 February 2011 (UTC)

And google's language tags are not reliable. Some of those 50 euro results are probably in other languages. If the EU insists on 50 cent (as a quantity, not an adjective) it is fighting idiom. Septentrionalis PMAnderson 19:23, 3 February 2011 (UTC)

Just like the BIPM insisting that a space separates the percent symbol and the numeric value (like 75 %): they’re being idealistic purists and are fighting an uphill battle against the entire inhabitants of this pale blue dot. Wikipedia follows the way the real world works. Greg L (talk) 04:19, 9 February 2011 (UTC)

The inhabitants of this pale blue dot speak a variety of languages and plurals differ, including amongst the various European nations using the Euro. However, this is the English-language Wikipedia, and it sounds odd to say "Euro are used by several nations as a common currency". --Pete (talk) 21:16, 15 February 2011 (UTC)

Perhaps because one should say "The euro is used by several nations as a common currency". −Woodstone (talk) 09:31, 17 February 2011 (UTC)

Massive automated delinking by User:Hmains

Note concerning the wording of the title: is there a risk that "Massive" is POV? Tony (talk) 05:47, 8 February 2011 (UTC)

The word accurately describes the actions of Hmains. Dolovis (talk) 05:05, 9 February 2011 (UTC)

I need a second opinion. User:Hmains is using the automated AWB tool to remove wikilinks to all “Century” articles (20th century, 19th century, 5th century BC, etc.[7],[8],[9] as recent examples of the delinking). I first became aware of his delinking here, which I reverted here with the edit comment “This is not an uncontroversial edit - possible misuse of AWB” which was ignored as Hmains's next edit was to restore his de-linking here. I am aware that there has been some debate about the linking of dates, but a “century” is not a date, and wikilinks to such “century” articles are useful navigational tools for our readers. He's edited/delinked thousands of articles, and he refused to stop when I warned him and tried to discuss it. [10], and Hmains' similar editing behaviour has also been previously discussed here, here, and here, but his use of AWB has not abated. I would appreciate your comments on whether or not Hmains use of AWB is contrary to its “Rules of Use” [11], and if so, can anything be done to stop his disruptive editing. Thank you for your attention. Dolovis (talk) 22:58, 5 February 2011 (UTC)

Thank heavens Hmain is doing the project a service like this. The community came out so heavily against the vague, low-value linking of chronological units in a huge RfC in 2009 that I wonder why anyone is launching a counter-offensive. The style guides reflect this, and also say to make links specific. Tony (talk) 01:12, 6 February 2011 (UTC)

Er, that was for year and date links. Century links are far different judging by the quality of the articles. This is not saying that every century term should be linked, but the 2009 RFC can't be directly applied here. --MASEM (t) 01:17, 6 February 2011 (UTC)

Concur with Tony. I see no 'useful navigational tool' in linking to a century article that is normally nothing more than a random list of unconnected events over a 100 year period. It would never even cross my mind to wikilink a century and I would always delete any such deprocated links that I tripped over. 21st CENTURY GREENSTUFF 01:20, 6 February 2011 (UTC)

• To be of best service to the reader, links should be specific, leading to topics which will aid the reader to a deeper understanding of the subject of the article. Generic links to centuries fails, IMHO, to impart such understanding. When the subject is Ashland, Alabama, which goes through the centuries just like any other locality, it is questionable indeed to divert the reader to 19th century or 21st century. --Ohconfucius ^¡digame! 02:44, 6 February 2011 (UTC)

• To User:Hmains: Please be sure to avoid operating your bot on Sewer cover in front of Greg L’s house. That particular article is supposed to have a totally obscene quantity of links. Yours truly, Greg L (talk) 03:05, 6 February 2011 (UTC)

P.S. (Oh… by the way…) Links are supposed to be germane and topical to the subject matter so they are something that has a half-way decent chance of being clicked upon by a reader studying “Ashland, Alabama.” The idea is not to link every fathomable word just because Wikipedia has an article on the subject and it can be linked to. In your differences, above, the subject matter of one of them was Ashland, Alabama, and one of the affected sentences was this:

---

Clay County was formed by an act of the Alabama General Assembly on December 7, 1866. Less than a year later, Ashland was established as the county seat on land donated by Hollingsworth Watts for the construction of a courthouse. Ashland was incorporated in 1871 and was named for 19th century statesman Henry Clay's Kentucky estate home.

---

How many of those links are really half-way well associated with Ashland? It’s interesting that our article, “19th century”, doesn’t mention “Clay County”—or even all of “Alabama” for that matter. The same goes for December 7 and for 1866. It’s safe to assume that someone reading up on “Ashland, Alabama” doesn’t need to be forked to an article that mentions the attack on Pearl Harbor. Normally, if one were to actually go read “1866”, one would *expect* to find that it has a bullet point mentioning how Clay County was formed that year—which would be pretty self-referential since the reader just read that much when the clicked the link. In this particular instance, the “1866” article doesn’t even mention that.

The linking principal to abide by is if it is a link that can help the reader better understand that particular subject and better prepare them for their future studies on the subject, then we provide a link. The reaction of the reader should be “Cool, it’s nice to know there is related reading on this subject!” Beyond that, we let the reader type things into the search field that are unrelated to Ashland, such as up and down, so as to not clutter up the article and turn it into a sea of mind-dumbing blue that obscures the truly valuable links that could assist with a better understanding of the study material at hand.

If the desire is to just ensure that readers are knowledgeable that “19th century” exists for further reading on the whole tangential subject of “old”, adding it to the See also section is a better way to accomplish that. Greg L (talk) 03:35, 6 February 2011 (UTC)

If they had deleted "Athletes of the 20th century" (which would be relevant), or they were putting "World conflict I" then you might have had a case - If you read the MilHist MoS you can clearly see that all references to that war are written "World War I"

Secondly I would have thought that starting discussions here first would have been the correct way to go.

Thirdly have you considered removing the links to [[1871]], [[1899]], and [[1930s]] yourself? That would negate the need for anyone else doing it for you and so stop any chance of a problem arising in the first place.

(I have also mentioned this on the MilHist talk for clarification that the ENGVAR violation is still a current issue) Chaosdruid (talk) 06:44, 6 February 2011 (UTC)

I am not trying to suggest that every incident of 13th century and 5th century BC should be linked, but the massive automated delinking of all Century links by User:Hmains will result in these articles being orphans. My concern is not the delinking over-linked articles. My concern is that such delinking should be done in a careful way to avoid mistakes, and it appears to me that [Hmains is not being selective or careful when using AWB to perform such delinkings on the massive scale as he is continuing to do. His misuse of AWB results in some controversial edits, which is an abuse of AWB. I am asking him to slow down, and to stop using the automated tools for such delinking edits. Dolovis (talk) 14:44, 6 February 2011 (UTC)

People scratched their heads for examples of these chronological items that were relevant and useful enough to bother linking. The numbers are tiny, and those that should remain should be piped to something specific, and better still, link to something more specific than a century. That will avoid the bot. Let's remember than chronological articles all link massively to each other, so will never be orphans. And then there's the main-page exposure they get. I'm not feeling sorry for them such that the emotive word "orphan" would suggest. Tony (talk) 15:00, 6 February 2011 (UTC)

I disagree with this statement, at least in as all-encompassing as it is for years or the like. Centuries are de facto "eras" of recorded human knowledge, on part with terms like "Dark ages" and "Industrial era". Certainly some links to these are effectively the same as if linking years ("SomePerson lived during the 4th and 5th Century", are bad links), but I would think when we start talking about events and the like that are attributed to specific centuries, links back to those centuries can be useful. Of course, this is based on the presumption that the century articles are written out like 20th century and less like 19th century.

But the point is this: while I can agree we should never link years or more specific dates, and that the majority of century links aren't likely necessary, there's more cases where century or even decade links in non-calender articles can be more useful and germane, and automatic delinking of these is not warranted. --MASEM (t) 15:20, 6 February 2011 (UTC)

• Let’s keep focused, can we? The broad principal for linking is to link only those articles that have a fairly reasonable prayer of being of value in understanding the subject matter at hand in its broad—but reasonably focused—context. A few years ago, we had some editors who, (apparently after shouting “Down, down clear writing” somewhere in Egypt), came here under the banner of “Build the Web®™©”. That slogan clearly was interpreted as a mandate to cross-link every conceivable word and phrase merely because articles existed to which one could generate a link. The result was articles that were seas of blue turd. Editors just need to use some common sense.

  If the subject matter is Ancient Egypt, I’m sure there are contexts where linking to a century makes sense since the readers coming to that article are clearly interested in ancient civilization. Moreover, it’s fair to say that 3rd century BC actually mentions Egypt (three times, I see). As I stated above, if the article is Ashland, Alabama, linking to 19th century leads the reader to something that doesn’t mention all of Alabama once, let alone Ashland. We’re not here to “build the Web” by adding links merely because technology gives us the power to do so.

  I note this useless link in Ashland, Alabama: There were 854 households out of which 27.5% had children under the age of 18 living with them, 43.0% were married couples living together. Really? “Married couples”?? That resolves to our “Marriage” article. Now, that’s just stupid. Besides the fact that it is a link to totally generic information wholly unrelated to the topic of “Ashland, Alabama,” such a term is beyond-obvious to the type of readership coming to this article; we write for the middle of the bell curve of the intended or likely readership; if we wrote for 1st graders, we’d have a icon of of Barney the dinosaur at the top of all our articles with a little dialog balloon coming out of his mouth saying “Say kids, do you know where babies come from?” I’m deleting that link right now… Greg L (talk) 22:26, 6 February 2011 (UTC)

  • "Common knowledge" terms, most agree, shouldn't be linked every occurrence. But that said, they should be linked in immediately relevant topics; the term marriage probably should be linked in articles dealing with faith and marriage, divorce, homosexual couples, and so forth. We cannot flat-out say that these terms should never be linked. Same with the centuries. I agree they are common knowledge terms, but at the same time, they are historical eras. Articles that deal with events that are linked to these centuries or eras should not omit the links to these just because they are common. There's also need of encouragement to link to articles that are structured like a hypothetical "Wars of the 14th century" instead of the actual century article if one can be that more specific. And that comes back to the problem: it is not the value of the links here, but the use of automated tools to delink them should not be used since the value of the link and possible replacement requires human evaluation. If the article starts being processed through quality control towards FA, then the relevance and appropriateness of each link can be pointed out as being excessive or not, but all those steps are human controlled.
  • Or, more to wit: overlinking in non-GA/FA articles harms nothing - it is a problem, but so is proper reference formatting, consistency of sections, copyediting, image use, and so on. They are problems dealt with by human editors in time. The use of bots to seemingly address the apparent immediacy of the issue is overexaggerating the problem. I can see it being done with years only because there's less human input, but centuries or other types of links are less obvious. If it must be done with bots or automated tools (I don't believe is the case), there then must be a way to block a bot/tool at the per-page or per-link level from making the change based on the editors' consensus on the page in question, otherwise, we are going to be right back at Arbcom like with the date-delinking issue from 2 years ago. --MASEM (t) 03:31, 7 February 2011 (UTC)
    • Yeah, I pretty much agree with your premiss of your first bullet point. Of course “Marriage” needs to be linked if it is in a relevant article. The above example wasn’t relevant (not even close) so I de‑linked it. As to your second bullet point (bots), I’ll stay out of that public square while the anti-bot dudes come riding through the on horses and camels whipping the poor slobs down below. Greg L (talk) 03:35, 7 February 2011 (UTC)
      • WP:AWB is "semi-automated"; it is not an automated WP:bot. That is, AWB allows HMains, or any user, to manually review each edit before accepting it. If HMains were to use AWB to unlink "marriage" in the Ashland article, and didn't unlink it in a faith and marriage article, then he would be using AWB appropriately. Potential unlinkings that haven't been done are an argument against automated bots, but not against AWB which is semi-automated. HMains' edit to "Ancient maritime history" is much more questionable than the hypothetical edits he didn't perform. Art LaPella (talk) 03:57, 7 February 2011 (UTC)

        • 'Marriage' is certainly overlinked. I come across it time and again in articles about places, in sections about demographics, where it is almost systematically linked to married couples. This is just an abuse of the wikilinking as 'marriage' would never conceivably have any role in deepening the understanding of any single article about a location, unless perhaps it was the place with the lowest or highest marriage rate in the world, or some such fact which the marriage article would go in depth to discuss. I'm not hooked on AWB, and would be much less inclined to use it if not for the endemic overlinking which pervades this encyclopaedia. Ohconfucius ^¡digame! 00:55, 8 February 2011 (UTC)

        • (Oh…) Well, then this is just a minor edit dispute that doesn’t even belong here. It belongs only on the talk pages of User:Dolovis and User:HMains. I’m done here on this one, I find it to be a bunch of “Waaah”. User: Dolovis certain has received his or her “second opinion.” Greg L (talk) 04:02, 7 February 2011 (UTC)
          • Yes, AWB is semi-automated. But I've used it, and when working on my own articles, I know how easy it is to think "Oh, I trust all changes AWB makes, I'll let it do it", and bam, something bad happened (which I was then able to correct). And agree that the specific issue here is between the AWB-using editor, and the one that wants the links in the article. But the general issue comparing the century links to year links, is where my concern is. I fully agree with how dates were delinked then for years - and thus reasonable that AWB can be included to trim them - but people are jumping to the conclusion that century links were covered by the same decision, which is not true. And why editors using AWB or other semi-automated tools to link-strip need to be aware what they are actually doing. The fact that the AWB user here repeated the actions after being reverted suggests that possibly they weren't checking each edit manually. That's the start of the same issues that created the first problem. --MASEM (t) 04:14, 7 February 2011 (UTC)
            • If "the use of automated tools to delink them should not be used since the value of the link and possible replacement requires human evaluation", then AWB should not be used for almost any purpose, because nearly all AWB edits require human evaluation (that's why they aren't fully automated). But evidently you didn't mean that. Good, I can go back to my AWB. Art LaPella (talk) 00:33, 8 February 2011 (UTC)
• I don't think anything is solved. The complaining editor still thinks he is correct in asserting that xx century in articles should be linked and that I am 'abusing AWB' by having AWB assist in my removal of such purposeless, unjustifiable links. I expect he will just continue to complain and obstruct. Hmains (talk) 03:30, 8 February 2011 (UTC)

I'm sure Hmains operates AWB with care and skill. If only I could use it. When is the Windows-only mafia going to fix AWB so it can be used on a real computer? Tony (talk) 05:43, 8 February 2011 (UTC)

• Speaking of which, does anyone know whether, if I procure a copy of OEM WIndows 95, would it run on a Mac, or does it have to be a proper bought-off-the-shelf version? --Ohconfucius ^¡digame! 05:54, 8 February 2011 (UTC)

If the editor is reverting your changes, then maybe its best not to try to fix them. Again, there's a lot more time and energy to be spent on other matters when the evaluation of linking only really matters when the article is at FA/GA. --MASEM (t) 05:48, 8 February 2011 (UTC)

That's good advice (unless it's a high-traffic article and the reversion patently silly). There are plenty more fields to plough. I do encourage bot-operators to engage with reverters, though; it's important at least to communicate with those who have issues. Tony (talk) 05:51, 8 February 2011 (UTC)

The Rf's were not just for single dates - they also included centuries as far as I remember. The summary is here Chaosdruid (talk) 05:52, 8 February 2011 (UTC)

Checking the result of that (just to verify), its clear that centuries and other date-type things have a lot more link allowances than plain year/date links, which even more so points to the fact that unless you're sure it's a wasted link, fighting over their delinking is a waste of time. Ultimately, FA/GA is where every link and lack of link should be validated, and nowhere else is the time and effort to fight that really valuable. Only in specific cases that we've been IDing over the last few months (common geography terms, bare dates and years, etc.) does it make sense to ensure that they are properly removed if not part of the accepted uses. --MASEM (t) 06:00, 8 February 2011 (UTC)

I briefly watch articles after I use AWB. I choose many of those articles because they are heavily read, not because they are FA/GA, and I have often communicated with reverters. My user space even has pre-written pages for that purpose. Of course communication doesn't have to be aggressive enough to make me the object of an investigation like this one. In particular, re-reverting doesn't communicate much that can't be said in words. Art LaPella (talk) 06:17, 8 February 2011 (UTC)

(ec) And conversely why not do it when AWB opens the page and makes other changes, it is just a tiny fraction of a second longer to load, two extra seconds to check it and it's done. Are you seriously suggesting leaving rubbish lying around until an article goes to GA/FA? I understand "if its not broken..." but if you are there and you have your tools out and it is not going to take more than a few seconds extra, why not ? Chaosdruid (talk) 06:19, 8 February 2011 (UTC)

Is there even a mote of this discussion thread that pertains to changing the guidelines of MOSNUM? Greg L (talk) 04:08, 9 February 2011 (UTC)

This discussion was brought here because Hmains has been using AWB in violation of its Rules of Use. I am seeking a consensus that will ask him to slow down, and to stop using the automated tools for making so many quick and hurried delinking edits. Dolovis (talk) 05:16, 9 February 2011 (UTC)

Could you explicitly refer to this "rule" and explain how it is being breached? Tony (talk) 11:47, 9 February 2011 (UTC)

I too would be interested in this explaination, as the removal of these links has been requested by the community through a very large Rfc and as such is not a controversial use of AWB. -DJSasso (talk) 14:12, 9 February 2011 (UTC)

See Rules of Use for AWB:

Rule#2: Don't edit too quickly - Hmains is making thousands of edits with AWB, often making hundreds of edits in an hours with 10 or more edits a minute. (See Hmains' edits history)

Rule #3: Don't do anything controversial with it - In addition to my ignored warning of controversial use here, Hmains has also been warned here, and has been the subject of other discussions concerning his misuse of AWB.

Rule #4: Avoid making insignificant or inconsequential edits - Most of Hmains edits are of a trivial nature. Dolovis (talk) 16:57, 9 February 2011 (UTC)

After studying this, this, and this, I conclude that the consensus interprets Rule 4 as applying to edits that don't change the appearance of a page, not to removing a visible link. Art LaPella (talk) 02:55, 10 February 2011 (UTC)

So even if he has not broken Rule 4, Hmains has still broken Rules 2 and 3. Dolovis (talk) 06:16, 10 February 2011 (UTC)

• Dolovis, I can't help sensing that you're extremely angry. You're certainly assuming bad faith in Hmains (whom I've had almost nothing to do with, so I can't be accused of bias on that count). Just because you don't agree with the community's decision in 2009 not to link chronological units doesn't mean you should be conducting a vendetta against this editor. I think these "rules" you cite are open to interpretation. IMO, Hmains hasn't broken any rules at all. Tony (talk) 10:35, 10 February 2011 (UTC)

I haven't read the whole thread but that's my opinion:

• Edits don't violate rule number 4. They are not trivial.
• Since there was a big discussion in unlinking dates (which I agree) there should be some warning by Hmains to some page before going and massively unlinking. The reason is that maybe there were some nice ideas of which pages to avoid etc. Something like the thing we do now in the wrong place.
• Since there are many edits a bot approval should be asked even if the edits were manual and supervised. BAG tries to supervise this kind of edits even if they aren't strictly done by bots (automated processes, etc.) -- Magioladitis (talk) 10:50, 10 February 2011 (UTC)

The pages to avoid are those that are explicitly on chronological topics: year articles, century articles, etc. I trust that Hmains has white-listed these. Um ... I'm pretty sure it's not the number of edits, but their speed, that influences how interested BAG is. Even then, I've come to learn that there's no hard-and-fast guideline about this. Now, why are we still discussing the linking of centuries in normal articles? The community has said "no". Several times, actually. Century articles are widely linked into a strong network of chronological articles (although I can't say they're all strong articles, sadly). In addition, one features on the main page of WP every day. Can't ask for more than that. Tony (talk) 13:11, 10 February 2011 (UTC)

No, there is no assurance that the community has said no to this; this is your assumption, the same type assumption that started the previous problems. The RFC/Arbcom case was about years and day/month links, nothing else, and strictly because of the ties to autoformatting. You cannot assume it applies to centuries or other day periods. It may be that the community is of your opinion on this, but you have no measuring stick to know that (and no, just because the regulars here agree does not mean it extends to everyone else). I personally see more value in the century link at certain times than bare year and day links, and recognize that most instances of century terms likely shouldn't be linked, but nowhere close to the scrubbing and removal that has been done for year and day. --MASEM (t) 14:06, 10 February 2011 (UTC)

I won't hold a mirror up to your rhetorical bolding technique. That decades, centuries and millennia should not generally be linked has been common practice, and as far as I see was the assumption in the community's judgements on the matter of linking chronological terms. Your attempts to wikilawyer are not working: even you go along with it independently, above: "while I can agree we should never link years or more specific dates, and that the majority of century links aren't likely necessary", although you provide your own opinion that there are cases "where century or even decade links in non-calender articles can be more useful and germane, and automatic delinking of these is not warranted". And just why would we want to privilege vague link-targets over specific years, months, days, and dates? The style guides are insistent that links should be focused. Tony (talk) 14:26, 10 February 2011 (UTC)

You're missing the point I'm saying. You are saying consensus exists when there is no evidence of such. You may be right, but you or others cannot enforce that without showing that exists. Without that evidence there's no consensus to edit-war (particularly with semi-auto tools like AWB) to remove those links. I may personally agree with their removal more of the time but not at the same frequency as years or day links, but I strongly disagree with the vigilante approach that regulars here seem to be suggesting to use for their removal. --MASEM (t) 15:06, 10 February 2011 (UTC)

Units of Measurement

Wikipedia is an international institution and as such should conform to international conventions as much as possible. This is particularly true for the English language Wikipedia, which is read by users from all over the world. This means that units of measures should prefer the SI system over any other nonstandard or regional system. Legacy units (e.g. feet, miles, cubits, leagues, etc) are fine and may be used in certain contexts, but absolutely not without expressing the values in conventional units as well. Most people in this world have no idea of how many stones they weigh or how many square feet their house measures, so please let's not be provincial and let's allow people from other countries to understand what we're talking about. Cheers —Preceding unsigned comment added by 81.100.22.178 (talk) 22:11, 9 February 2011 (UTC)

How is this different from existing guidance? Titoxd^{(?!? - cool stuff)} 22:28, 9 February 2011 (UTC)

MOSNUM’s current guidelines are the product of ten-thousand debates on this issue and strike the proper balance, where the primary unit of measure depends on a variety of issues, including—but certainly not limited to—whether there is a strong national tie to a particular article (e.g. American football will be primarily in yards) as well as what units of measure are used by the sources that are cited in the article (e.g. distances to other galaxies are often in light years, not petameters). Plus, MOSNUM requires that appropriate conversions are made available. If you, I.P. 81.100.22.178, saw an article that is not satisfactory, it may well not be in compliance with MOSNUM and needs to be improved. You are welcome to do so. Just please be sure to read and understand the applicable sections of MOSNUM before correcting the articles. Greg L (talk) 04:41, 10 February 2011 (UTC)

Community input invited on a request for bot approval to automatically delink dates

A request for bot approval has been filed for a task that will, among other things, "delinks full dates (but not lone day-month strings or years), days, months, decades, centuries", "removes direct links to full dates, whether ISO8601, dd mmm yyyy or mmm dd, yyyy, including piped links of same to chronological articles in almost any imaginable form" (per WP:UNLINKDATES) and ensure articles uses a consistent date format throughout.

A member of the Bot Approvals Group has requested community input to determine if community consensus exists for an automated process of this nature.

Editors are invited to comment on the feasibility and desirability of the automated task here. –xeno^talk 16:23, 15 February 2011 (UTC)

Point of order on dates

Recently, User:Dl2000 changed [12] the section on date consistency withing articles and references.

Before

• Dates in article body text should all have the same format.
• Dates in article references should all have the same format.

These requirements apply to dates in general prose and reference citations, but not to dates in quotations or titles.

After

Dates in the article's body text and references should all have the same format. These requirements do not apply to dates in in quotations or in titles of works.

At priori, this seems like a minimal change, but as far as I know the bullets were intentionally separate, as articles text could read "25 September 2007" while references use "25 Sep 2007". The intention of the bullets were to prevent prose mixing (Julia ate a poisoned apple on 25 June 2005. She died three days later on June 28.) and reference mixing (Jones, J. (20 September 2008)... Smith, J. (March 20, 2005)., but to allow for differences between citation and prose.

I reverted the edit and gave more detailed (and hopefully clear) explanations, along with examples, for the current version. I also added the thing about accessdates, which had somehow been omitted. If things changed since the last date-delinking drama, (or that I'm simply wrong about some things), we should have an RFC on it. Otherwise, I doubt anyone will have much success in making anything stick, and it'll be a drama-fest all over again.

Headbomb {talk / contribs / physics / books} 20:30, 15 February 2011 (UTC)

The thing about dates in references is that they are not part of written text, and we can therefore use the ISO format of YYYY-MM-DD, which has no national ties and is widely used in Asia. I think it best if all dates (apart from quotations) in an article have the same format, but sometimes there are non-text items, such as tables and infoboxes where ISO is the standard. It's all a bit of a muddle, really, and I think no good can come out of trying to make everything the same. --Pete (talk) 21:08, 15 February 2011 (UTC)

I strenuously object to calling the format YYYY-MM-DD the ISO format. The ISO 8601 format is not suitable for Wikipedia articles because it requires all dates to be expressed in the Gregorian calendar. Wikipedia sometimes uses sources that were written before the adoption of the Gregorian calendar in the country of publication. Jc3s5h (talk) 21:16, 15 February 2011 (UTC)

I'd rather see YYYY-MM-DD avoided and dates throughout the article to be consistant but with allowances for abbreviations (e.g. 12 Sep 2003). JIMp _talk·cont 21:30, 15 February 2011 (UTC)

Jc3s5h, I explained my edits and their reason. Some people agreed that the YYYY-MM-DD format was fine in references/accessdates, others disagreed, but they were neither banned nor deprecated. Removing this from the section will have bots and people edit war on the format of accessdates. If this changed since the date-delinking drama, or you claim that YYYY-MM-DD is in fact banned and/or depreated, hold the RFC, or link to one establishing exactly that, rather than revert war and create drama. Headbomb {talk / contribs / physics / books} 21:55, 15 February 2011 (UTC)

The status quo, as I understand it before the failed proposal Wikipedia:Mosnum/proposal on YYYY-MM-DD numerical dates, was that YYYY-MM-DD format dates were allowed when space was limited, and that date formats should be consistent. It could be, and often was, argued, that space should be conserved in citations and thus the YYYY-MM-DD format was legitimate there, provided all the dates in the references section were in the Gregorian calendar. However, no consensus has ever been achieved about whether it is legitimate to use a date such as February 15, 2011 for the publication date of a newspaper, while using the YYYY-MM-DD format for the accessdate parameter. Jc3s5h (talk) 22:05, 15 February 2011 (UTC) struck out Jc3s5h (talk) 22:16, 15 February 2011 (UTC)

Or phrased differently, there was no consensus that it is illegitimate to use February 15, 2011 for the publication date, and YYYY-MM-DD format for the accessdate. What is not forbidden is allowed. Headbomb {talk / contribs / physics / books} 22:09, 15 February 2011 (UTC)

On review of versions of the MOSNUM before the Wikipedia:Mosnum/proposal on YYYY-MM-DD numerical dates I see that the examples do indeed show a mixture of non-all-numeric publication dates with YYYY-MM-DD access dates, so I withdraw my objection to Headbomb's change. Jc3s5h (talk) 22:16, 15 February 2011 (UTC)

• As I've said before, those gobbledy telephone-number dates are bad bad bad for our readership, wherever they appear. I'd be banning them altogether. Tony (talk) 23:35, 15 February 2011 (UTC)

• This isn’t intended as a personal attack and it flat isn’t. But the protestation of User:Jc3s5h (I strenuously object to calling the format YYYY-MM-DD the ISO format) belies the simple fact that it is. Jc3s5h’s activity on the “ISO 8601” article betrays his or her interest in the format.

  Wikipedia does best when the writing style chosen for a given article does not draw undo attention to itself. No matter how well intentioned the ISO’s proposal is to ensure that another “Y2K bug” never happens, “2011-02-17T00:32Z” is a writing style that does draw attention to itself and causes *!* brain-interrupts that interfere with the transparent communication of thought.

  There is no one-size-fits-all solution because articles closely associated with European issues will read most naturally in one format whereas still other articles have different needs. Maritime-related articles directed to an expert maritime readership might be best written in Zulu time; I don’t know and I don’t care. Astronomy and other specialty subjects should be left to the specialists of those articles and the denizens of WP:MOSNUM shouldn’t have to become 15-minute experts in everything. We should simply have some broad-brushed, global principals that ensures each article uses the most natural, human-readable dates that best serve the likely readership.

  I agree with Tony: Setting aside special circumstances such as tables, where space is at a premium, articles should generally express dates with the month written out in English; lose the telephone numbers.

  I’m also concerned about this text: Dates in article body text should all have the same format. That is overly prescriptive. Once it has been written that something occurred at a Boston Red Sox game on “February 13, 2011” or that so-n-so was beheaded in France on “2 February, 1799”, there is no need to keep repeating the year (I’m addressing the prescription for the “*all* have the same format” here) if the text in the next sentence says “and they beheaded his wife on “7 March”; the year is clear enough without belaboring the text with more numerals (although the ISO would be displeased because such an expression would cause problems with data exchange if you tried to buy a plane ticket with such sketchy information).

  We’re here to write fluid, most-natural-reading, clear prose; not promote some Star Trek-style star‑date format or some standard organization’s all-numeric expression of temporal measures (for corpuscular beings caught in linear time in this universe). Greg L (talk) 01:04, 17 February 2011 (UTC)

How about all dates in day-before-month format or month-before-day format including variations on each (no day, no year, abbreviated month names, etc.)? No telephone-number dates. JIMp _talk·cont 01:31, 17 February 2011 (UTC)

Good, Jimp: I’m pleased to see that we might have a developing consensus to bury, once and for all, dates that look like ham radio frequencies. Can you provide some examples of what date expressions would be prescribed and those that would be proscribed? Greg L (talk) 01:34, 17 February 2011 (UTC)

I'd group formats like this:

• Group A: day only or no day at all:
  • the 7th or 7 (day only)
  • September, Sep or Sept. (month only)
  • 2011 (year only)
  • September 2011, Sep 2011 or Sept. 2011 (month and year)
• Group B: day before month
  • 7 September 2011, 7 Sep 2011 or 7 Sept. 2011 (day, month and year)
  • 7 September, 7 Sep or 7 Sept. (day and month)
• Group C: month before day
  • September 7, 2011, Sep 7, 2011 or Sept. 7, 2011 (month, day and year)
  • September 7, Sep 7 or Sept. 7 (month and day)

plus various adaptations of these for ranges. Don't combine formats from Groups B and C on the one article. If you use the "Sep" style abbreviation, stick to it. Same for the "Sept." style (what was that called again?). JIMp _talk·cont 07:04, 17 February 2011 (UTC)

• Sorry, does this mandate "7th September" and such? Please no. Tony (talk) 07:47, 17 February 2011 (UTC)

I still don't see any reason to ban or deprecated the YYYY-MM-DD format in footnotes, tables, infoboxes, accessdates, etc.. As much as some very vocal people hate them with a passion, consensus did not favour deprecating them last time anyone bothered asking the community for their feelings on the issue, and I don't see what has changed since then. Headbomb {talk / contribs / physics / books} 08:01, 17 February 2011 (UTC)

Fully agree seeing no reason to ban or deprecate the YYYY-MM-DD date format. No consensus about such a measure was ever reached, and I do not see it approaching here. −Woodstone (talk) 09:25, 17 February 2011 (UTC)

No, "7th September" would be out. "The 7th" would be used with no month e.g. "They made camp on 5 September but were attacked by lions on the 7th." JIMp _talk·cont 13:56, 17 February 2011 (UTC)

superscripts on ordinals with a variable in them

I'm not too thrilled in general with the guideline not to superscript "th", "rd", "st", but surely it's completely wrong in cases like n^th, which is just about unreadable without the superscript. The workaround n-th strikes me as nonstandard. I think we should at least put in an exception to the guide for this case. --Trovatore (talk) 00:33, 17 February 2011 (UTC)

• In cases like n^th, ignore MOSNUM then. MOSNUM can’t be all-seeing / all-knowing and have answers for everything. Wikipedia:Ignore all rules (If a rule prevents you from improving or maintaining Wikipedia, ignore it) was written for a reason. Greg L (talk) 01:11, 17 February 2011 (UTC)

What about "If a rule prevents you from improving or maintaining Wikipedia, change it."? JIMp _talk·cont 01:23, 17 February 2011 (UTC)

I certainly agree that the MoS should not try to cover everything. However, this was brought to my attention by an edit at Kolmogorov complexity where someone changed n^th to n-th based on WP:ORDINAL. Constructions like n^th are common enough that I would expect the issue to recur unless clarified.

In fact I'm not certain I have an ideal solution. n^th looks less good to me on the screen than I had expected. But I definitely don't like nth. The problem (or at least a problem) with n-th is that the hyphen looks like a minus sign. I think n^th is least bad, but maybe not by as much as I had first thought. --Trovatore (talk) 01:24, 17 February 2011 (UTC)

• (responding to Jimp:) If it’s not too much of a holy-war subject like date formats or rewriting the Koran, changing MOSNUM can be a solution too. Often, too much time is wasted here on WT:MOSNUM doing the Wikipedia-equivalent of Turkish-prison butt-stabbings.

  On more than one occasion, I’ve had editors do a drive-by on an article I was shepherding and doing all the heavy lifting on (translation: I acted like *owned* it). After I reverted them, they claimed “But… MOS or MOSNUM said this ‘n’ that.” If the rule amounted to trying to cram a square peg into a round hole, my response was “Yeah, and the rule is retarded, here’s why, so I’m ignoring it.” It helps if you actually have a clue of what you’re talking about if you take a stand like that, but it can at least be therapeutic in that crazy world that is the collaborative writing environment of Wikipedia.

  But it’s nice, Jimp, that you are willing to devote the time necessary to actually get anything accomplished here. Greg L (talk) 01:31, 17 February 2011 (UTC)

n^th dimension, etc: that's a good example of where "common sense" should be used WRT the style guides. But all the same, the exception might be best written in.

It's an interesting issue, though. What do you do with, say, $(n+1)^{\mathrm {st} }$? As written there it looks like "en plus one to the power st". But $(n+1)-\mathrm {st}$ looks like "en plus one minus st", and $(n+1)\mathrm {st}$ looks like "en plus one times st". To say nothing of the barbarians who prefer $(n+1)^{\mathrm {th} }$. Honestly it would be nice to get a good answer to these; I don't have one myself. --Trovatore (talk) 09:30, 17 February 2011 (UTC)

If you put the hyphen outside the math tags, it doesn't get converted to a minus sign: $(n+1)$-st.